T
Tom
I am trying to use some of the statistical functions on a vector of numbers
that contains occasional alphabetic data, specifically the string "FALSE".
Average(), Stdev(), Avedev(), and Median() seem to work fine, although I am
suspicious that they are not necessarily ignoring the alpha data, but
substituting a value, probably 0, for the missing data. So my first two
questions follow ...
1. Are they substituting values or are they just ignoring the alpha data;
i.e. giving the correct results?
2. If they *are* substituting values, is there anything I can do, short of
writing a script to remove the offending data, to give me an accurate
result?
Additionally, I am using the Percentile() worksheet function. That returns
a #NUM! value. I assume that it works differently than the functions I
mentioned before. My third question follows ...
3. Is there anything I can do, short of writing a script to remove the
offending data, to give me an accurate result from Percentile() with the
alpha data in the vector?
I would appreciate anybody's thoughts on this.
Thanks,
Tom
that contains occasional alphabetic data, specifically the string "FALSE".
Average(), Stdev(), Avedev(), and Median() seem to work fine, although I am
suspicious that they are not necessarily ignoring the alpha data, but
substituting a value, probably 0, for the missing data. So my first two
questions follow ...
1. Are they substituting values or are they just ignoring the alpha data;
i.e. giving the correct results?
2. If they *are* substituting values, is there anything I can do, short of
writing a script to remove the offending data, to give me an accurate
result?
Additionally, I am using the Percentile() worksheet function. That returns
a #NUM! value. I assume that it works differently than the functions I
mentioned before. My third question follows ...
3. Is there anything I can do, short of writing a script to remove the
offending data, to give me an accurate result from Percentile() with the
alpha data in the vector?
I would appreciate anybody's thoughts on this.
Thanks,
Tom